NoSQL Data Models - LMU

DATABASESYSTEMSGROUP

NoSQL Data Models

The 4 Main NoSQL Data Models:

• Key/Value Stores

• Document Stores

• Wide Column Stores

• Graph Databases

Big Data Management and Analytics 71


NoSQL Data Models

Key/Value Stores:

• Most simple form of database systems

• Store key/value pairs and retrieve values by keys

• Values can be of arbitrary format


10213

10334

10023

51


NoSQL Data Models

Key/Value Stores:

• Consistency models range from Eventual consistency toserializibility

• Some systems support ordering of keys, which enablesefficient querying, like range queries

• Some systems support in-memory data maintenance, someuse disks

→ There are very heterogeneous systems



NoSQL Data Models

Key/Value Stores - Redis:

• In-memory data structure store with built-in replication, transactions and different levels of on-disk persistence

• Support of complex types like lists, sets, hashes, …

• Support of many atomic operations

>> SET val 1>> GET val => 1>> INCR val => 2>> LPUSH my_list a (=> ‘a‘)>> LPUSH my_list b (=> ‘b‘,‘a‘)>> RPUSH my_list c (=> ‘b‘,‘a‘,‘c‘)>> LRANGE my_list 0 1 => b,aBig Data Management and Analytics 74


NoSQL Data Models

Key/Value Stores – The Redis cluster model:

• Data is automatically sharded across nodes

• Some degree of availability, achieved by master-slave architecture (but cluster stops in the event of larger failures)

• Easily extendable



NoSQL Data Models


Big Data Management and Analytics

A

B

C

Hash slotsNodes

0

5000

5001

10000

10001

14522

A

B

C

Hash slotsNodes

0

4000

4001

8000

8001

12000

D12001

14522

add

node

remove

node

A

B

Hash slotsNodes

0

7500

7501

14522


NoSQL Data Models


Big Data Management and Analytics

A

B

C

MasterNodes Hash slots

0

5000

5001

10000

10001

14522

Hash slots 5001 – 10000 cannot be used anymore

A

B

C

MasterNodes Hash slots

0

5000

5001

10000

10001

14522

A‘

B‘

C‘

Slave Nodes

ReplicatedHash slots

0

5000

5001

10000

10001

14522

Slave node B‘ is promoted asthe new master and hash slots5001 – 10000 are still available


Big Picture

CAP Theorem:


C

A PA

CACID

BASE

All clients alwayshave the same view

of the data

The system works welldespite physicalnetwork partitions

Each client can al-ways read and write

AP-Systems

AC-Systems CP-Systems- Redis

- Dynamo

Key/Value Stores

- RDBMSs (MySQL, Postgres, …)


NoSQL Data Models

Document Stores:

• Store documents in form of XML or JSON

• Semi-structured data records that do not have a homogeneous structure

• Columns can have more than one value (arrays)

• Documents include internal structure, or metadata

• Data structure enables efficient use of indexes



NoSQL Data Models

Document Stores:

Given following text:

→ Find all <contact>s where <zip> is “12345“


Max MustermannMusterstraße 12D-12345 Musterstadt

<contact><first_name>Max</first_name><last_name>Mustermann</last_name><street>Musterstraße 12</street><city>Musterstadt</city><zip>12345</zip><country>D</country>

</contact>


NoSQL Data Models

Document Stores:

• Data stored as documents in binary representation (BSON)

• Similarly structured documents are bundled in collections

• Provides own ad-hoc query language

• Supports ACID transactions on document level



NoSQL Data Models

Document Stores:

MongoDB Data Management:– Automatic data sharding– Automatic re-balancing

• Multiple sharding policies:– Hash-based sharding: Documents are distributed

according to an MD5 hash → uniform distribution– Range-based sharding: Documents with shard key values

close to one another are likely to be co-located on the same shard → works well for range queries

– Location-based sharding: Documents are partitioned wrtto a user-specified configuration that associates shard key ranges with specific shards and hardware



NoSQL Data Models

Document Stores:

MongoDB Consistency & Availabilty:

• Default: Strong consistency (but configurable)

• Increased availability through replication

– Replica sets consist of one primaryand multiple secondary members

– MongoDB applies writes on theprimary and then records the operations on the primary’s oplog



Big Picture

CAP Theorem:


C

A PA

CACID

BASE


of the data



AP-Systems


- Dynamo

Key/Value StoresDocument Stores

- MongoDB

- CouchDB



NoSQL Data Models

Wide Column Stores:

• Rows are identified by keys

• Rows can have different numbers of columns (up tomillions)

• Order of rows depend on key values (locality is important!)

• Multiple rows can be summarized to families (or tablets)

• Multiple families can be summarized to a key space



NoSQL Data Models

Wide Column Stores:


Key Space

Column Family

RowKey

Column Name

Value

Column Name

Value

Column Name

Value

RowKey

Column Name

Value

RowKey

Column Name

Value

Column Name

Value

Column Name

Value

Column Name

Value

Column

Column Family


NoSQL Data Models

Wide Column Stores:


Key Space „Edibles“

Column Family „Fruit“

Applecolor

„green“

weight

95

variety

„Granny Smith“

Cherrycolor

„red“

Lemoncolor

„yellow“

weight

50

origin

„Egypt“

flavor

„sour“

Column Family „Vegetable“

Carrot2015-08-11

65

2015-08-12

50

…

…

2015-09-21

87


NoSQL Data Models

Wide Column Stores:

• Developed by Facebook, Apache project since 2009

• Cluster Architecture:– P2P system (ordered as rings)– Each node plays the same role

(decentralized)– Each node accepts read/write operations

• User access through nodes via Cassandra Query Language (CQL)


N

NN

N

NN

N

NN

NN

N

NN

N

N


NoSQL Data Models

Wide Column Stores:

Consistency

• Tunable Data Consistency (choosable per operation)

• Read repair: if stale data is read, Cassandra issues a readrepair → find most up-to-date data and update stale data

• Generally: Eventually consistent

• Main focus on availability!



Big Picture

CAP Theorem:


C

A PA

CACID

BASE


of the data



AP-Systems


- Dynamo

Key/Value StoresDocument StoresWide Column Stores

- MongoDB

- CouchDB

- Cassandra

- HBase



NoSQL Data Models

Graph Databases:

• Use graphs to store and represent relationships betweenentities

• Composed of nodes and edges

• Each node and each edge can contain properties (Property-Graphs)


AliceAlice

BobBob

CarolCarolDaveDave

lentmoney to

knows


NoSQL Data Models

Graph Databases:

Alice is a friend of Bob and vice versa. They both love the movie „Titanic“.


name = „Alice“ name = „Bob“

title = „Titanic“


NoSQL Data Models

Graph Databases:





Person Person

Movie


NoSQL Data Models

Graph Databases:





Person Person

Movie

is a friend of

is a friend of

loves loves


NoSQL Data Models

Graph Databases:

• Master-Slave Replication (no partitioning!)

• Consistency: Eventual Consistency (tunable to Immediate Consistency)

• Support of ACID Transactions

• Cypher Query Language

• Schema-optional



Big Picture

CAP Theorem:


C

A PA

CACID

BASE


of the data



AP-Systems

AC-Systems CP-Systems


- Redis

- Dynamo

Key/Value StoresDocument StoresWide Column StoresGraph Databases

- MongoDB

- CouchDB

- Cassandra

- HBase

- Neo4J